xref: /dokuwiki/inc/pageutils.php (revision b7ab6f2f26a5380cb3d0aaeb64116aa79b9dcebe) 
1<?php
2/**
3 * Utilities for handling pagenames
4 *
5 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
6 * @author     Andreas Gohr <andi@splitbrain.org>
7 * @todo       Combine similar functions like {wiki,media,meta}FN()
8 */
9
10/**
11 * Fetch the an ID from request
12 *
13 * Uses either standard $_REQUEST variable or extracts it from
14 * the full request URI when userewrite is set to 2
15 *
16 * For $param='id' $conf['start'] is returned if no id was found.
17 * If the second parameter is true (default) the ID is cleaned.
18 *
19 * @author Andreas Gohr <andi@splitbrain.org>
20 */
21function getID($param='id',$clean=true){
22  global $conf;
23
24  $id = isset($_REQUEST[$param]) ? $_REQUEST[$param] : null;
25
26  //construct page id from request URI
27  if(empty($id) && $conf['userewrite'] == 2){
28    //get the script URL
29    if($conf['basedir']){
30      $relpath = '';
31      if($param != 'id') {
32        $relpath = 'lib/exe/';
33      }
34      $script = $conf['basedir'].$relpath.basename($_SERVER['SCRIPT_FILENAME']);
35    }elseif($_SERVER['DOCUMENT_ROOT'] && $_SERVER['SCRIPT_FILENAME']){
36      $script = preg_replace ('/^'.preg_quote($_SERVER['DOCUMENT_ROOT'],'/').'/','',
37                              $_SERVER['SCRIPT_FILENAME']);
38      $script = '/'.$script;
39    }else{
40      $script = $_SERVER['SCRIPT_NAME'];
41    }
42
43    //clean script and request (fixes a windows problem)
44    $script  = preg_replace('/\/\/+/','/',$script);
45    $request = preg_replace('/\/\/+/','/',$_SERVER['REQUEST_URI']);
46
47    //remove script URL and Querystring to gain the id
48    if(preg_match('/^'.preg_quote($script,'/').'(.*)/',$request, $match)){
49      $id = preg_replace ('/\?.*/','',$match[1]);
50    }
51    $id = urldecode($id);
52    //strip leading slashes
53    $id = preg_replace('!^/+!','',$id);
54  }
55
56  // Namespace autolinking from URL
57  if(substr($id,-1) == ':' || ($conf['useslash'] && substr($id,-1) == '/')){
58    if(@file_exists(wikiFN($id.$conf['start']))){
59      // start page inside namespace
60      $id = $id.$conf['start'];
61    }elseif(@file_exists(wikiFN($id.noNS(cleanID($id))))){
62      // page named like the NS inside the NS
63      $id = $id.noNS(cleanID($id));
64    }elseif(@file_exists(wikiFN($id))){
65      // page like namespace exists
66      $id = $id;
67    }else{
68      // fall back to default
69      $id = $id.$conf['start'];
70    }
71    header("Location: ".wl($id,'',true));
72  }
73
74  if($clean) $id = cleanID($id);
75  if(empty($id) && $param=='id') $id = $conf['start'];
76
77  return $id;
78}
79
80/**
81 * Remove unwanted chars from ID
82 *
83 * Cleans a given ID to only use allowed characters. Accented characters are
84 * converted to unaccented ones
85 *
86 * @author Andreas Gohr <andi@splitbrain.org>
87 * @param  string  $raw_id    The pageid to clean
88 * @param  boolean $ascii     Force ASCII
89 */
90function cleanID($raw_id,$ascii=false){
91  global $conf;
92  global $lang;
93  static $sepcharpat = null;
94
95  global $cache_cleanid;
96  $cache = & $cache_cleanid;
97
98  // check if it's already in the memory cache
99  if (isset($cache[$raw_id])) {
100    return $cache[$raw_id];
101    }
102
103  $sepchar = $conf['sepchar'];
104  if($sepcharpat == null) // build string only once to save clock cycles
105    $sepcharpat = '#\\'.$sepchar.'+#';
106
107  $id = trim($raw_id);
108  $id = utf8_strtolower($id);
109
110  //alternative namespace seperator
111  $id = strtr($id,';',':');
112  if($conf['useslash']){
113    $id = strtr($id,'/',':');
114  }else{
115    $id = strtr($id,'/',$sepchar);
116  }
117
118  if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
119  if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
120
121  //remove specials
122  $id = utf8_stripspecials($id,$sepchar,'\*');
123
124  if($ascii) $id = utf8_strip($id);
125
126  //clean up
127  $id = preg_replace($sepcharpat,$sepchar,$id);
128  $id = preg_replace('#:+#',':',$id);
129  $id = trim($id,':._-');
130  $id = preg_replace('#:[:\._\-]+#',':',$id);
131
132  $cache[$raw_id] = $id;
133  return($id);
134}
135
136/**
137 * Return namespacepart of a wiki ID
138 *
139 * @author Andreas Gohr <andi@splitbrain.org>
140 */
141function getNS($id){
142  $pos = strrpos($id,':');
143  if($pos!==false){
144    return substr($id,0,$pos);
145  }
146  return false;
147}
148
149/**
150 * Returns the ID without the namespace
151 *
152 * @author Andreas Gohr <andi@splitbrain.org>
153 */
154function noNS($id) {
155  $pos = strrpos($id, ':');
156  if ($pos!==false) {
157    return substr($id, $pos+1);
158  } else {
159    return $id;
160  }
161}
162
163/**
164* Returns the current namespace
165*
166* @author Nathan Fritz <fritzn@crown.edu>
167*/
168function curNS($id) {
169    return noNS(getNS($id));
170}
171
172/**
173* Returns the ID without the namespace or current namespace for 'start' pages
174*
175* @author Nathan Fritz <fritzn@crown.edu>
176*/
177function noNSorNS($id) {
178    global $conf;
179
180    $p = noNS($id);
181    if ($p == $conf['start']) {
182        $p = curNS($id);
183        if ($p == false) {
184            return noNS($id);
185        }
186    }
187    return $p;
188}
189
190/**
191 * returns the full path to the datafile specified by ID and
192 * optional revision
193 *
194 * The filename is URL encoded to protect Unicode chars
195 *
196 * @author Andreas Gohr <andi@splitbrain.org>
197 */
198function wikiFN($raw_id,$rev='',$clean=true){
199  global $conf;
200
201  global $cache_wikifn;
202  $cache = & $cache_wikifn;
203
204  if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
205    return $cache[$raw_id][$rev];
206  }
207
208  $id = $raw_id;
209
210  if ($clean) $id = cleanID($id);
211  $id = str_replace(':','/',$id);
212  if(empty($rev)){
213    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
214  }else{
215    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
216    if($conf['compression']){
217      //test for extensions here, we want to read both compressions
218       if (@file_exists($fn . '.gz')){
219          $fn .= '.gz';
220       }else if(@file_exists($fn . '.bz2')){
221          $fn .= '.bz2';
222       }else{
223          //file doesnt exist yet, so we take the configured extension
224          $fn .= '.' . $conf['compression'];
225       }
226    }
227  }
228
229  if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
230  $cache[$raw_id][$rev] = $fn;
231  return $fn;
232}
233
234/**
235 * Returns the full path to the file for locking the page while editing.
236 *
237 * @author Ben Coburn <btcoburn@silicodon.net>
238 */
239function wikiLockFN($id) {
240  global $conf;
241  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
242}
243
244
245/**
246 * returns the full path to the meta file specified by ID and extension
247 *
248 * The filename is URL encoded to protect Unicode chars
249 *
250 * @author Steven Danz <steven-danz@kc.rr.com>
251 */
252function metaFN($id,$ext){
253  global $conf;
254  $id = cleanID($id);
255  $id = str_replace(':','/',$id);
256  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
257  return $fn;
258}
259
260/**
261 * returns an array of full paths to all metafiles of a given ID
262 *
263 * @author Esther Brunner <esther@kaffeehaus.ch>
264 */
265function metaFiles($id){
266   $name   = noNS($id);
267   $dir    = metaFN(getNS($id),'');
268   $files  = array();
269
270   $dh = @opendir($dir);
271   if(!$dh) return $files;
272   while(($file = readdir($dh)) !== false){
273     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
274       $files[] = $dir.$file;
275   }
276   closedir($dh);
277
278   return $files;
279}
280
281/**
282 * returns the full path to the mediafile specified by ID
283 *
284 * The filename is URL encoded to protect Unicode chars
285 *
286 * @author Andreas Gohr <andi@splitbrain.org>
287 */
288function mediaFN($id){
289  global $conf;
290  $id = cleanID($id);
291  $id = str_replace(':','/',$id);
292    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
293  return $fn;
294}
295
296/**
297 * Returns the full filepath to a localized textfile if local
298 * version isn't found the english one is returned
299 *
300 * @author Andreas Gohr <andi@splitbrain.org>
301 */
302function localeFN($id){
303  global $conf;
304  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
305  if(!@file_exists($file)){
306    //fall back to english
307    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
308  }
309  return $file;
310}
311
312/**
313 * Resolve relative paths in IDs
314 *
315 * Do not call directly use resolve_mediaid or resolve_pageid
316 * instead
317 *
318 * Partyly based on a cleanPath function found at
319 * http://www.php.net/manual/en/function.realpath.php#57016
320 *
321 * @author <bart at mediawave dot nl>
322 */
323function resolve_id($ns,$id,$clean=true){
324  // if the id starts with a dot we need to handle the
325  // relative stuff
326  if($id{0} == '.'){
327    // normalize initial dots without a colon
328    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
329    // prepend the current namespace
330    $id = $ns.':'.$id;
331
332    // cleanup relatives
333    $result = array();
334    $pathA  = explode(':', $id);
335    if (!$pathA[0]) $result[] = '';
336    foreach ($pathA AS $key => $dir) {
337      if ($dir == '..') {
338        if (end($result) == '..') {
339          $result[] = '..';
340        } elseif (!array_pop($result)) {
341          $result[] = '..';
342        }
343      } elseif ($dir && $dir != '.') {
344        $result[] = $dir;
345      }
346    }
347    if (!end($pathA)) $result[] = '';
348    $id = implode(':', $result);
349  }elseif($ns !== false && strpos($id,':') === false){
350    //if link contains no namespace. add current namespace (if any)
351    $id = $ns.':'.$id;
352  }
353
354  if($clean) $id = cleanID($id);
355  return $id;
356}
357
358/**
359 * Returns a full media id
360 *
361 * @author Andreas Gohr <andi@splitbrain.org>
362 */
363function resolve_mediaid($ns,&$page,&$exists){
364  $page   = resolve_id($ns,$page);
365  $file   = mediaFN($page);
366  $exists = @file_exists($file);
367}
368
369/**
370 * Returns a full page id
371 *
372 * @author Andreas Gohr <andi@splitbrain.org>
373 */
374function resolve_pageid($ns,&$page,&$exists){
375  global $conf;
376  $exists = false;
377
378  //keep hashlink if exists then clean both parts
379  if (strpos($page,'#')) {
380    list($page,$hash) = split('#',$page,2);
381  } else {
382    $hash = '';
383  }
384  $hash = cleanID($hash);
385  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
386
387  // get filename (calls clean itself)
388  $file = wikiFN($page);
389
390  // if ends with colon or slash we have a namespace link
391  if(substr($page,-1) == ':' || ($conf['useslash'] && substr($page,-1) == '/')){
392    if(@file_exists(wikiFN($page.$conf['start']))){
393      // start page inside namespace
394      $page = $page.$conf['start'];
395      $exists = true;
396    }elseif(@file_exists(wikiFN($page.noNS(cleanID($page))))){
397      // page named like the NS inside the NS
398      $page = $page.noNS(cleanID($page));
399      $exists = true;
400    }elseif(@file_exists(wikiFN($page))){
401      // page like namespace exists
402      $page = $page;
403      $exists = true;
404    }else{
405      // fall back to default
406      $page = $page.$conf['start'];
407    }
408  }else{
409    //check alternative plural/nonplural form
410    if(!@file_exists($file)){
411      if( $conf['autoplural'] ){
412        if(substr($page,-1) == 's'){
413          $try = substr($page,0,-1);
414        }else{
415          $try = $page.'s';
416        }
417        if(@file_exists(wikiFN($try))){
418          $page   = $try;
419          $exists = true;
420        }
421      }
422    }else{
423      $exists = true;
424    }
425  }
426
427  // now make sure we have a clean page
428  $page = cleanID($page);
429
430  //add hash if any
431  if(!empty($hash)) $page .= '#'.$hash;
432}
433
434/**
435 * Returns the name of a cachefile from given data
436 *
437 * The needed directory is created by this function!
438 *
439 * @author Andreas Gohr <andi@splitbrain.org>
440 *
441 * @param string $data  This data is used to create a unique md5 name
442 * @param string $ext   This is appended to the filename if given
443 * @return string       The filename of the cachefile
444 */
445function getCacheName($data,$ext=''){
446  global $conf;
447  $md5  = md5($data);
448  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
449  io_makeFileDir($file);
450  return $file;
451}
452
453/**
454 * Checks a pageid against $conf['hidepages']
455 *
456 * @author Andreas Gohr <gohr@cosmocode.de>
457 */
458function isHiddenPage($id){
459  global $conf;
460  if(empty($conf['hidepages'])) return false;
461
462  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
463    return true;
464  }
465  return false;
466}
467
468/**
469 * Reverse of isHiddenPage
470 *
471 * @author Andreas Gohr <gohr@cosmocode.de>
472 */
473function isVisiblePage($id){
474  return !isHiddenPage($id);
475}
476
477/**
478 * Checks and sets HTTP headers for conditional HTTP requests
479 *
480 * @author   Simon Willison <swillison@gmail.com>
481 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
482 * @param    timestamp $timestamp lastmodified time of the cache file
483 * @returns  void or void with previously header() commands executed
484 */
485function http_conditionalRequest($timestamp){
486  // A PHP implementation of conditional get, see
487  //   http://fishbowl.pastiche.org/archives/001132.html
488  $last_modified = substr(gmdate('r', $timestamp), 0, -5).'GMT';
489  $etag = '"'.md5($last_modified).'"';
490  // Send the headers
491  header("Last-Modified: $last_modified");
492  header("ETag: $etag");
493  // See if the client has provided the required headers
494  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
495    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
496  }else{
497    $if_modified_since = false;
498  }
499
500  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
501    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
502  }else{
503    $if_none_match = false;
504  }
505
506  if (!$if_modified_since && !$if_none_match){
507    return;
508  }
509
510  // At least one of the headers is there - check them
511  if ($if_none_match && $if_none_match != $etag) {
512    return; // etag is there but doesn't match
513  }
514
515  if ($if_modified_since && $if_modified_since != $last_modified) {
516    return; // if-modified-since is there but doesn't match
517  }
518
519  // Nothing has changed since their last request - serve a 304 and exit
520  header('HTTP/1.0 304 Not Modified');
521  exit;
522}
523
524//Setup VIM: ex: et ts=2 enc=utf-8 :
525